RSTP and switch reboot

[Ron Cole_1]

Hello all,

We are trying to track down a problem with some devices locking up whenever a switch reboots. One question raised by my engineering staff is how RSTP "starts up" when a switch is rebooted. Is there ever any chance of a small "broadcast storm" when a switch with RSTP first comes online? They are concerned that the switch needs to relearn the RSTP paths and a loop is created until the paths are "relearned". I don't think it's true since the port settings are written to memory and most likely applied during the POST sequence, but I need something concrete to back that up.

Thanks,

Ron Cole
IT Manager, IED Inc.
rcole@iedaudio.com

[Matt Hobbs]

It is somewhat possible with the default spanning-tree enabled configuration as it puts the ports edge-port status to 'yes' by default with RSTP.

This means that a port immediately starts forwarding and only if it detects a loop will it start STP blocking that port.

To prevent this from happening, you should set all of your switch-to-switch links as 'no' edge-port.

no spanning-tree 49-50 edge-port

I think on the newer firmware releases it is now:

no spanning-tree 49-50 auto-edge-port

You can then verify the edge-port setting in the 'show span' output.

[Ron Cole_1]

I just double checked, and all switch-to-switch links have Edge set to "No". Is there anything else I need to look for?